Matches bioRxiv preprints with their corresponding PubMed paper.
First, install docker for hosting the PostgreSQL database. Then, the following python libraries must be installed via pip:
numpy
numba
tqdm
lxml
psycopg2-binary
fasttext
requests
preprint-match was tested with python 3.7, but it may work with other python3 versions.
First, run docker compose up -d in the home directory of this repository to start a postgres instance in docker. Then, a local copy of PubMed must be created. First, run cd data; python3 download.py to download PubMed files one-by-one through the FTP service, and then preprocess the results and save them in the database with python load_pubmed.py in the home directory. These steps must only be performed once before matching papers, and then periodically after that if desired to update the database.
To match a preprint and paper, use the API provided in match_from_doi.py. It contains a function, match_from_doi, that takes a list of preprint DOIs and returns a list of matching paper DOIs. Example usage:
from match_from_doi import match_from_doi
print(match_from_doi(['preprintdoi1', 'preprintdoi2']))
Contact petereckmann(at)gmail(dot)com for help using the tool or for access to the data.